I have a batch of new PCBs with ENIG (electroless nickel / immersion gold) surface finish. My assembly house has had two unsuccessful attempts at populating the boards. We're trying to troubleshoot the situation.
The assembly house is suggesting that it is Black Pad [1] [2], which is caused during PCB manufacturing but isn't exposed until assembly/soldering. This would, of course, shift the responsibility away from the assembly process.
With ENIG, the nickel is the bonding surface whereas the gold is simply there to protect the nickle. During reflow, the gold dissolves into the solder. You don't know there is "black pad" until the underlaying nickel is exposed.
In my case, we have a 15-up array but we are only populating a few of the individual boards. The remaining boards do not have solder applied to them; all that happens to them is that they go through the oven twice.
We are seeing something that looks like black pad on these post-oven boards:
My question: Will the gold get burned away somehow when exposed to reflow conditions, even if there is no solder applied to the pads? Or might this be an entirely different problem?
Additional info: The image makes it appear that if might be soldermask on the pads, but in real life you can see that it isn't. Here's the best image my equipment will give me. You can (kind of) see that the discoloration is different that the intentionally-placed soldermask.
Another edit. I was able to get a better image, shown below. I think I know what's happening, but I don't want to influence any upcoming answers.